1. Field of the Invention
The present invention relates to a method for controlling the signal integration time of a signal integration type image sensing device having a plurality of signal integration type radiation-sensitive elements.
2. Description of the Prior Art
In a signal integration type image sensing device such as a CCD (Charge Coupled Device) image sensor, a BBD (Bucket Brigade Device) image sensor, or an MOS-image sensor (Self-Scan Photodiode Array), it is necessary to automatically control the signal integration time, that is, the time for integrating the electric signals (charge) which are generated in correspondence to the intensity of incident light, in order to enlarge the dynamic range of the radiation-electric (charge) conversion.
As a method for controlling the signal integration time of a signal integration type image sensing device of this type, there is known a conventional method in which it is discriminated if the output level of the image sensing device during readout is suitable, too high or too low. If the output level of the image sensing device is too high, the signal integration time for the subsequent integration operations is controlled so that the signal integration times may become shorter in a stepped manner. On the other hand, if the output level from the image sensing device is too low, the signal integration times for the subsequent signal integration operations are so controlled that the signal integration times may be lengthened in a stepped manner.
However, with such a conventional method, control must be performed so that a plurality of mutually different discrete signal integration times are preset and a suitable one is selected in each signal integration operation. Therefore, the configuration of a control circuit for such control becomes complex. Furthermore, since the change in time is made in a stepped manner, the control is unable to follow abrupt changes in the intensity of incident radiation. When the intensity of incident radiation periodically changes (e.g., due to effects of artificial illumination), the output level fluctuates upon every signal integration operation even with the same signal integration time, if the scanning device of the image sensing device is not well synchronized with the change in the intensity of incident radiation. This method is also poor in reliability since the control involved in this method is not the so-called "real time" control.
In consideration of this, the assignee of the present application proposed, in Japanese Patent Application No. 54-17425 (Japanese Laid-Open Patent Application No. 55-110477), a method for controlling the signal integration time of a signal integration type image sensing device wherein a timing for initiating readout of integration signals from all radiation-sensitive elements up to this point is determined by detecting if the integrated value of the signal from at least one of these radiation-sensitive elements reaches a predetermined value after initiation of signal integration. This method is novel and is advantageous in that control of the signal integration time may be performed in "real time" and in a nonstepped manner. Therefore, this method solved various problems with the conventional method.
In the embodiment disclosed in the specification and the drawings of this Japanese Patent Application No. 54-17425, in order to detect if the integrated value of the signal from at least one of the radiation-sensitive elements reaches a predetermined value after initiation of signal integration, an overflow of a signal generated from the radiation-sensitive element through an overflow drain gate is detected, the overflow drain gate being incorporated for the radiation-sensitive element for the purpose of preventing blooming. However, with this method, the overflowed signal generated at the radiation-sensitive element is clipped, and a video signal (image scanning signal) from the image sensing device is adversely affected.